[alsa-devel] Fwd: [LAD] kernel 2.6.29 rc6 no working rme multiface 1

[Takashi Iwai]

At Tue, 3 Mar 2009 01:54:14 +0300,
Dmitry Baikov wrote:
> 
> On Mon, Mar 2, 2009 at 7:50 PM, Takashi Iwai <tiwai at suse.de> wrote:
> > At Mon, 2 Mar 2009 06:59:35 +0300,
> > Dmitry Baikov wrote:
> >>
> >> ---------- Forwarded message ----------
> >> From: Dmitry Baikov <dsbaikov at gmail.com>
> >> Date: Mon, Mar 2, 2009 at 5:49 AM
> >> Subject: Re: [LAD] kernel 2.6.29 rc6 no working rme multiface 1
> >> To: "m.wolkstein at gmx.de" <m.wolkstein at gmx.de>
> >> Cc: The Linux Audio Developers' Mailing List
> >> <linux-audio-dev at lists.linuxaudio.org>
> >>
> >>
> >> On Mon, Mar 2, 2009 at 4:59 AM, m.wolkstein at gmx.de <m.wolkstein at gmx.de> wrote:
> >> > hiho,
> >> > my name is michael wolkstein and i wrote the first time to this list.
> >> > sorry about my english. but i am no english native-speaker.
> >> >
> >> > after booting kernel 2.6.29 rc6 i get a problem with my rme multiface 1 connected true pcmcia cardbusinterface.
> >> >
> >> > i get this message after insert my card.
> >> > RME Hammerfall DSP 0000:16:00.0: enabling device (0000 -> 0002)
> >> > RME Hammerfall DSP 0000:16:00.0: PCI INT A -> GSI 16 (level, low) -> IRQ 16
> >> > Hammerfall-DSP: no Digiface or Multiface connected!
> >> > RME Hammerfall DSP 0000:16:00.0: PCI INT A disabled
> >> > RME Hammerfall DSP: probe of 0000:16:00.0 failed with error -5
> >> >
> >> > i have compile the kernel included alsa module.
> >> > also i have installed the newest alsa-firmware 1.0.19 and newest alsa-tools 1.0.19.
> >> >
> >> > my old kernel 2.6.24 load both firmwares correct.
> >> >
> >> > the old one 1.0.16 (debian unstable)
> >> > and the new 1.0.19 download from the alsa-project page
> >> >
> >> > the new kernel produce errors. kernel module snd_hdsp is already loaded with the new kernel.
> >>
> >>
> >> Have the same problem here.
> >> 2.6.24 works, 2.6.28 doesn't (tried 2.6.8-ARCH and 2.6.28.7-1-ARCH).
> >> I reported the problem to alsa-devel list, but they said I have hw
> >> problems (I think I do not).
> >
> > Could you check the latest sound git tree to be sure?
> >
> >   git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6.git
> 
> Does not oops.
> But the card does not work either.
> 
> > Also the problem might be rather the cardbus side...
> I suspect, too.
> Because it works in my laptop. But fails on a desktop with pci-cardbus adapter.

That explains.  A PCI-board is often problematic...

> >
> >> The latest arch kernel gives oops, if the card is inserted before booting.
> >> It crashes during udev event processing (i.e. something from userspace
> >> triggers the problem)
> >> If anybody is really interested, I can try to provide kernel backtrace
> >> (either type it in, or try to use serial console).
> >
> > Oops is bad, and yes, it'd be helpful if you can show the backtrace.
> 
> No serial console :( My new M/B have only pin header and no rs232
> jack. Need to find/build an adapter.
> So. what I see (without stack addresses and some offsets):

Looks like an accessing to the uninitialized device.
Does the patch below fix?

We should initialize spinlock & co at the very beginning, though...


Takashi

---
diff --git a/sound/pci/rme9652/hdsp.c b/sound/pci/rme9652/hdsp.c
index dc65fe1..314e735 100644
--- a/sound/pci/rme9652/hdsp.c
+++ b/sound/pci/rme9652/hdsp.c
@@ -3740,6 +3740,9 @@ static irqreturn_t snd_hdsp_interrupt(int irq, void *dev_id)
 	midi0status = hdsp_read (hdsp, HDSP_midiStatusIn0) & 0xff;
 	midi1status = hdsp_read (hdsp, HDSP_midiStatusIn1) & 0xff;
 
+	if (!(hdsp->state & HDSP_InitializationComplete))
+		return IRQ_HANDLED;
+
 	if (audio) {
 		if (hdsp->capture_substream)
 			snd_pcm_period_elapsed(hdsp->pcm->streams[SNDRV_PCM_STREAM_CAPTURE].substream);